Two Nvidia consumer GPUs, one Ada Lovelace and one Blackwell: 12 GB GDDR6X against 32 GB GDDR7 and 3.6x the memory bandwidth on the RTX 5090.
At a glance
Updated 5 minutes ago- Memory
- 12 GB
- GDDR6X
- Memory bandwidth
- 504 GB/s
- From
- $0.08 /GPU/hr
- 3 providers
- Memory
- 32 GB
- GDDR7
- Memory bandwidth
- 1.79 TB/s
- From
- $0.35 /GPU/hr
- 20 providers
Price comparison
Cheapest listed price per GPU per hour by billing type, and the median across providers.
| Billing | RTX 4070 | RTX 5090 | Difference |
|---|---|---|---|
| On-demand | $0.08 | $0.35 | 4.1x more expensive on RTX 5090 |
| Spot | $0.07 | $0.09 | 25% cheaper on RTX 4070 |
| Reserved | No reserved listings | $0.21 | |
| Median | $0.21 | $0.63 | 3x more expensive on RTX 5090 |
Median: Median across providers, one vote each, so a large catalog does not outweigh a small one.
Price history
RTX 4070 vs RTX 5090 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | RTX 4070 | RTX 5090 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 12 GB GDDR6X | 32 GB GDDR7 | 2.7x more on RTX 5090 |
| Bandwidth | 504 GB/s | 1,792 GB/s | 3.6x higher on RTX 5090 |
| Bus width | 192-bit | 512-bit | 2.7x wider on RTX 5090 |
| Compute · dense | |||
| FP4 | Not supported | 1,676 TFLOPS | RTX 5090 only |
| FP8 | 116.6 TFLOPS | 419 TFLOPS | 3.6x higher on RTX 5090 |
| FP16 / BF16 | 58.3 TFLOPS | 209.5 TFLOPS | 3.6x higher on RTX 5090 |
| INT8 | 233.2 TOPS | 838 TOPS | 3.6x higher on RTX 5090 |
| FP32 | 29.1 TFLOPS | 104.8 TFLOPS | 3.6x higher on RTX 5090 |
| Platform | |||
| Interconnect | PCIe only | PCIe only | |
| Board power | 200 W | 575 W | 2.9x higher on RTX 5090 |
| Released | Q2 2023 | Q1 2025 | 2 years newer on RTX 5090 |
| Precision | |||
| Supported formats | FP8FP16BF16TF32FP32INT4INT8 | FP4FP8FP16BF16TF32FP32INT8 | |
Compute: Vendor peak figures, dense and per GPU. Not measured throughput.
Source: RTX 4070 datasheet and RTX 5090 datasheet.
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | RTX 4070 /GPU/hr | RTX 5090 /GPU/hr |
|---|---|---|
Vast.ai
|
On-Demand $0.08 | On-Demand $0.39 |
Salad
|
On-Demand $0.22 | On-Demand $0.45 |
Theta EdgeCloud
|
On-Demand $0.17 | -- |
HyperAI
|
-- | On-Demand $0.35 |
GPU.ai
|
-- | On-Demand $0.44 |
GPUhub
|
-- | On-Demand $0.46 |
Lium
|
-- | On-Demand $0.50 |
Nova Cloud
|
-- | On-Demand $0.63 |
Runpod
|
-- | On-Demand $0.69 |
3 providers list the RTX 4070 and 20 the RTX 5090. Every listing is on the RTX 4070 page and the RTX 5090 page.
Similar GPUs
Other comparisons in the same class.
Our data for Nvidia RTX 4070 was last updated on Sept. 7, 2026, and for Nvidia RTX 5090 on Sept. 7, 2026.